Manufacture cell scheduling (MCS) is a NP-hard problem, and the traditional optimization methods have their limitations when dealing with MCS. Immune algorithm that simulates the biologic immunity system is an intelligent optimization method and has potential to solve MCS. Multi-colony diploid immune algorithm (MDIA) is put forward to solve MCS. Firstly, the model of MCS is set up. Then, a typical benchmark problem is simulated with MDIA and other algorithms. Finally, simulation results show that MDIA is both effective and superior to other algorithms.